机构地区: 韶关学院计算机科学学院
出 处: 《韶关学院学报》 2006年第12期31-34,共4页
摘 要: 在超分辨率重构图像中,可以观察到四种不同的重构噪声,分别是边缘振荡效应、梳毛效应、边缘锯齿效应和方块效应.本文提出了一种能同时有效地滤除这四种重构噪声的边界自适应滤波算法.实验结果表明,该算法能在保持图像细节信息的同时较好地滤除各种重构噪声. In the super- resolution reconstructed images, four types of reconstruction noise can De observed, namely, ringing effect, combing effect, zigzag effect, and blocking effect. In this paper, a novel edge adaptive filter was proposed to suppress all these noises. Experiments demonstrated the effectiveness of the proposed algorithm.